A possible definition of life.

Tom Roger Kise tomk at ifi.uio.no
Mon May 5 09:09:27 EST 1997


Bill House wrote:
> 
> Matthew Stanfield <mattst at cogs.susx.ac.uk> wrote in article
> <5j2io1$an6 at infa.central.susx.ac.uk>...
> >
> > "Life (on Earth) consists of all things built by DNA."
> >
> 
> What is it about DNA that qualifies it over, say carbon, as the chemical
> compound whose properties define life?  If some other entity exhibited those
> properties, would not it also be considered alive?
> 
> IMO, life is not a physical compound, but an emergent phenomenon of complex
> entities that allows them to replicate, transform energy to locally reduce
> entropy, and to evolve new behaviors and phenotypes that enhance their ability
> to replicate and transform energy. This probably goes against convention, but I
> happen to think convention is to specific to biological systems.

I partly agree with your last sentence here. Someone has also pointed
out the mix between biological question about life and the metaphysical
question about life in this discussion. If not necessarely biological, I
think this thread really is not about the metaphysical part (try
comp.ai.philosophy), but you will all decide yourself.

If we go back to the beginning of it all, what created the first object
alive? (whatever being 'alive' is). Many agree that life started somehow
in the primal soup, a 'soup' of molecules constantly mixing with
eachother. The energy from all this came from the physical forces (like
magnetic), the heat from the inner of earth and the sun. If this is so,
_when_ did actually 'life' itself occur for the first time? Did life
gradually evolve from some primitive replicators or did it happen that
some quite unlikely lucky mix of molecules together made a primitive,
but in molecular terms large organism which had the ability to
reproduce? If the last answer is correct, okey, define the
characteristics of this organism -> you will have the definition of
life. If life gradually evolved, we will not be able to say when it
occured and then certainly not define that organisms characteristics
(although we know a lot of what these _might_ be). The question is quite
similar to the question 'when did the first organism of a specie occur?'
As we know evolution, an organism of that sort never occured, the
evolution is in these matters gradual. In an evolutionary line within
related species we can take one individual from the beginning of the
line and one from the end of the line, and we have no trouble defining
which species they belong to. But what about the individuals between?
When did that crucial change occur? What this brings up, is that with a
gradually evolution of life, there will be in between forms which is
hard to recognize as life or not.

I strongly believe that the discussion of what life is (the biological
question that is), always depend on everyones personal experience,
everyones intuition. A computerprogram is not alive, a rabbit is, no
doubth! There are lifeforms on earth which need scientists to 'become
alive', our intuition tells us that that corrals are dead stone.

One thing which I at this point do believe is needed in all sorts of
life which should appeal to our intuiton is this: when reproducing, any
lifeform must make a _physical_ baby (or egg or whatever). The parent
must _build_ a child from parts taken from its environment (whith itself
as the fabric/tool/machine). Computerprograms which definitly can hold a
lot of the characteristics which we find important about life, is
therefore not alive. Computerprograms live _inside_ physical matter,
they are not _built from it. You can certainly object with this
scenario: a computer with robotic arms, which can bulid a copy of
itself. Apart from arguments I have heard earlier of a lifeform not
being dependent on an external energysource (which can be contradicted
by symbiotic lifeforms perhaps), I have no good answer to that one.

-- 
Tom Kise (tomk at ifi.uio.no)
Casparisgt. 11 (http://www.ifi.uio.no/~tomk)
0174 OSLO (22 11 23 65)



More information about the Mol-evol mailing list